Lightning Mapping Array Data Research Articles

Investigating Temporal Characteristics of Polarimetric and Electrical Signatures in Three Severe Storms: Insights from the VORTEX-Southeast Field Campaign

Abstract The dual-polarization radar characteristics of severe storms are commonly used as indicators to estimate the size and intensity of deep convective updrafts. In this study, we track rapid fluctuations in updraft intensity and size by objectively identifying polarimetric fingerprints such as ZDR and KDP columns, which serve as proxies for mixed-phase updraft strength. We quantify the volume of ZDR and KDP columns to evaluate their utility in diagnosing temporal variability in lightning flash characteristics. Specifically, we analyze three severe storms that developed in environments with low-to-moderate instability and strong 0–6-km wind shear in northern Alabama during the 2016–17 VORTEX-Southeast field campaign. In these three cases (a tornadic supercell embedded in stratiform precipitation, a nontornadic supercell, and a supercell embedded within a quasi-linear convective system), we find that the volume of the KDP columns exhibits a stronger correlation with the total flash rate. The higher covariability of the KDP column volume with the total flash rate suggests that the overall electrification and precipitation microphysics were dominated by cold cloud processes. The lower covariability with the ZDR column volume indicates the presence of nonsteady updrafts or a less prominent role of warm rain processes in graupel growth and subsequent electrification. Furthermore, we observe that the majority of cloud-to-ground (CG) lightning strikes a carried negative charge to the ground. In contrast to findings from a tornadic supercell over the Great Plains, lightning flash initiations in the Alabama storms primarily occurred outside the footprint of the ZDR and KDP column objects. Significance Statement This study quantifies the correlation between mixed-phase updraft intensity and total lightning flash rate in three severe storms in northern Alabama. In the absence of direct updraft velocity measurements, we use polarimetric signatures, such as ZDR and KDP columns, as proxies for updraft strength. Our analysis of polarimetric radar and lightning mapping array data reveals that the lightning flash rate is more highly correlated with the KDP column volume than with the ZDR column volume in all three storms examined. This contrasts with previous findings in storms over the central Great Plains, where the ZDR column volume showed higher covariability with flash rate. Interestingly, lightning initiation in the Alabama storms mainly occurred outside the ZDR and KDP column areas, contrary to previous findings.

Evaluation of Lightning Flash Rate Parameterizations in a Cloud‐Resolved WRF‐Chem Simulation of the 29–30 May 2012 Oklahoma Severe Supercell System Observed During DC3

AbstractEighteen lightning flash rate parameterization schemes (FRPSs) were investigated in a Weather Research and Forecasting model coupled with chemistry cloud‐resolved simulation of the 29–30 May 2012 supercell storm system observed during the Deep Convective Clouds and Chemistry (DC3) field campaign. Most of the observed storm's meteorological conditions were well represented when the model simulation included both convective damping and lightning data assimilation techniques. Newly‐developed FRPSs based on DC3 radar observations and Lightning Mapping Array data are implemented in the model, along with previously developed schemes from the literature. The schemes are based on relationships between lightning and various kinematic, structural, and microphysical thunderstorm characteristics (e.g., cloud top height, hydrometeors, reflectivity, and vertical velocity) available in the model. The results suggest the model‐simulated graupel and snow/ice hydrometeors require scaling factors to more closely represent proxy observations. The model‐simulated lightning flash trends and total flashes generated by each scheme over the simulation period are compared with observations from the central Oklahoma Lightning Mapping Array. For this supercell system, 13 of the 18 schemes overpredicted flashes by >100% with the group of FRPSs based on storm kinematics and structure (particularly updraft volume) performing slightly better than the hydrometeor‐based schemes. During the storm's first 4 hr, the upward cloud ice flux FRPS, which is based on the combination of vertical velocity and hydrometeors, well represents the observed total flashes and flash rate trend; while, the updraft volume scheme well represents the observed flash rate peak and subsequent sharp decline in flash rate.

Dancing Sprites Above a Lightning Mapping Array—An Analysis of the Storm and Flash/Sprite Developments

AbstractOne of the most enigmatic types of sprites is the dancing sprite which appears to dance above the storm as sequential luminous emissions a few 100 ms or less apart. Dancing sprites occur in relatively small proportion and many aspects of their generation remain unknown. We present a multi‐instrumental analysis of a 20‐hr duration Mesoscale Convective System (MCS) over the northwestern Mediterranean Sea on September 21, 2019, that produced 21 sprites recorded with a video camera, of which 19 (90) were dancing sprites. The asymmetric trailing stratiform MCS developed in strong convective conditions having a CAPE of 3,500 J . It formed several convective cores (up to 2,900 with cloud top temperature −C) and exhibited a bow echo structure during the sprite production period. Using Lightning Mapping Array data, we show that the sprite producing positive cloud‐to‐ground (SP + CG) flashes mainly initiated at the edge of the convective line on the side of the stratiform region. The flashes propagated 100–200 km across it, producing both positive and negative CG strokes. The 19 dancing sprite events included 49 sequences, of which 46 were associated with distinct SP + CG strokes and 3 with surges during the continuing current. An especially bright and wide sprite sequence was produced by three distinct SP + CG strokes that occurred within 3 ms and spread over 54 km. This sprite sequence could be classified as a new sprite category resembling a "wall" but structured in three groups, each associated with one of the +CG strokes.

Characterizing Charge Structure in Central Argentina Thunderstorms During RELAMPAGO Utilizing a New Charge Layer Polarity Identification Method.

A new automated method to retrieve charge layer polarity from flashes, named Chargepol, is presented in this paper. Using data from the NASA Lightning Mapping Array (LMA) deployed during the Remote sensing of Electrification, Lightning, And Mesoscale/microscale Processes with Adaptive Ground Observations (RELAMPAGO) field campaign in Cordoba, Argentina, from November 2018 to April 2019, this method estimates the polarity of vertical charge distributions and their altitudes and thicknesses (or vertical depth) using the very‐high frequency (VHF) source emissions detected by LMAs. When this method is applied to LMA data for extended periods of time, it is capable of inferring a storm's bulk electrical charge structure throughout its life cycle. This method reliably predicted the polarity of charge within which lightning flashes propagated and was validated in comparison to methods that require manual assignment of polarities via visual inspection of VHF lightning sources. Examples of normal and anomalous charge structures retrieved using Chargepol for storms in Central Argentina during RELAMPAGO are presented for the first time. Application of Chargepol to five months of LMA data in Central Argentina and several locations in the United States allowed for the characterization of the charge structure in these regions and for a reliable comparison using the same methodology. About 13.3% of Cordoba thunderstorms were defined by an anomalous charge structure, slightly higher than in Oklahoma (12.5%) and West Texas (11.1%), higher than Alabama (7.3%), and considerably lower than in Colorado (82.6%). Some of the Cordoba anomalous thunderstorms presented enhanced low‐level positive charge, a feature rarely if ever observed in Colorado thunderstorms.

Dart‐Leader and K‐Leader Velocity From Initiation Site to Termination Time‐Resolved With 3D Interferometry

AbstractSimultaneous data from two interferometers separated by 16 km and synchronized within 100 ns were collected for a thunderstorm near Langmuir Lab on October 23, 2018. Analysis via triangulation followed by a least squares fit to time of arrival across all six antennae produced a three‐dimensional interferometer (3DINTF) data set. Simultaneous Lightning Mapping Array data enabled an independent calculation of 3DINTF accuracy, yielding a median location uncertainty of 200 m. This is the most accurate verified result to date for a two‐station interferometer. The 3D data allowed profiling the velocity of multiple dart leaders and K leaders that followed the same channel. 3D velocities calculated from the in‐cloud initiation site to ground ranged from 3 × 106 to 20 × 106 m/s. Average velocity generally increased with subsequent leaders, consistent with increased conditioning of the channel. Also, all leaders showed a factor of 2–3 decrease in velocity as they proceeded over 15 km of channel. We speculate that the velocity decrease is consistent with energy lost in the reionization of the channel at the leader tip. This paper includes an appendix providing details of the triangulation technique used.

LMA observations of upward lightning flashes at the Säntis Tower initiated by nearby lightning activity

We present in this paper lightning current measurements, LMA (Lightning Mapping Array) data and fast antenna electric fields associated with upward flashes observed at the Säntis Tower during summer of 2017. The LMA network consists of six stations that were installed in the vicinity of the tower at distances ranging from 100 m to 11 km from it. Out of 20 LMA recorded flashes here we analyze in detail three so-called ‘other-triggered flashes’, triggered by preceding activity. Based on the lightning activity derived from the European Lightning Detection Network (EUCLID) in an area within 30 km from the tower and within a 1-s time window before the start of the upward tower flashes, only one out of 20 flashes was classified as ‘other-triggered’(OT). However, the investigations based on the LMA data reveal that 3 more flashes of the 20 analyzed were preceded by nearby activity and should therefore be classified as OT flashes. We analyze conditions conducive to the OT flashes, such as the charge structure of the clouds, polarity of preceding leaders and level of activity of the storm.The LMA source active time period was on average seven times higher for the OT flashes than that for self-initiated flashes.

CHAT: The Colorado Hail Accumulation from Thunderstorms Project

AbstractIn recent years, hail accumulations from thunderstorms have occurred frequently enough to catch the attention of the National Weather Service, the general public, and news agencies. Despite the extreme nature of these thunderstorms, no mechanism is currently in place to obtain adequate reports, measurements, or forecasts of accumulated hail depth. To better identify and forecast hail accumulations, the Colorado Hail Accumulation from Thunderstorms (CHAT) project was initiated in 2016 with the goals of collecting improved and more frequent hail depth reports on the ground as well as studying characteristics of storms that produce hail accumulations in Colorado. A desired outcome of this research is to identify predictors for hail-producing thunderstorms typically occurring along the Colorado Front Range that might be used as operational nowcast products in the future. During the 2016 convective season, we asked amateur meteorologists to send general information, photos, and videos on hail depth using social media. They submitted over 58 reports in Colorado with information on location, time, depth, and areal coverage of hail accumulations. We have analyzed dual-polarization radar and lightning mapping array data from 32 thunderstorms in Colorado, which produced between 0.5 and 50 cm of hail accumulation on the ground, to identify characteristics unique to storms with hail accumulations. This preliminary analysis shows how enhanced in-cloud hail presence and surface accumulation can be tracked throughout the lifetime of a thunderstorm using dual-polarization radar and lightning data, and how hail accumulation events are associated with large in-cloud ice water content, long hailfall duration, or a combination of these.

Properties of Negative Initial Leaders and Lightning Flash Size in a Cluster of Supercells

AbstractProperties of negative initial leaders (NILs) and flash size in a cluster of supercells with generally inverted charge structure in Oklahoma on 10–11 May 2010 are examined, primarily using Lightning Mapping Array data. A method to identify NILs from Lightning Mapping Array source is proposed and helps to reveal the multiple NILs properties and their distributions. The NILs in the supercell cluster have smaller speed (median 3‐D displacement speed: 0.65 × 105 m/s), relative to the previous reports in normal thunderstorms. Furthermore, median NIL speeds initially decrease with increasing height but begin increasing above 12 km. The NILs tend to decelerate during the early stage. The parameters characterizing flash duration and spatial size are also investigated. It is found that they all follow lognormal distributions and the spatial flash size is relatively small on average (median horizontal distance: 5.54 km). Most flashes (83.18%) extend primarily in the horizontal direction. Flash area shows an inverse relationship with flash density at their fast changes during storm evolution. Although large flash initiation density (FID) generally occurs in regions with small flash size, the smallest flash size is nearly not collocated with large FID value. In the regions with large FID, average flash duration roughly increases with increasing FID, while average flash area changes little. We proposed that the pattern of charge pockets and variation of charge density dominated by the strong kinematics are responsible for some new findings about the properties of NIL and flash size in the supercell cluster.

Microphysical and kinematic processes associated with anomalous charge structures in isolated convection.

Microphysical and kinematic characteristics of two storm populations, based on their macroscale charge structures, are investigated in an effort to increase our understanding of the processes that lead to anomalous (or inverted charge) structures. Nine normal polarity cases (mid-level negative charge) with dual-Doppler and polarimetric coverage that occurred in northern Alabama, and six anomalous polarity cases (mid-level positive charge) that occurred in northeastern Colorado are included in this study. The results show that even though anomalous polarity storms formed in environments with similar instability, they had significantly larger and stronger updrafts. Moreover, the anomalous polarity storms evidently have more robust mixed-phase microphysics, based on a variety of metrics. Anomalous polarity storms in Colorado have much higher cloud base heights and shallower warm cloud depths in this study, leading us to hypothesize that anomalous polarity storms have lower amounts of dilution and entrainment. We infer positively charged graupel, and therefore high supercooled water contents, in the mid-levels of the anomalous storms based on the relationship between colocations of graupel and inferred positive charge from Lightning Mapping Array data. Using representative updraft speeds and warm cloud depths, the time required for a parcel to traverse from cloud base to the freezing level was estimated for each storm observation. We suggest this metric is the key discriminator between the two storm populations and leads us to hypothesize that it strongly influences the amount of supercooled water and the probability of positive charge in the midlevels, leading to an anomalous charge structure.

Assimilation of Flash Extent Data in the Variational Framework at Convection-Allowing Scales: Proof-of-Concept and Evaluation for the Short-Term Forecast of the 24 May 2011 Tornado Outbreak

AbstractThis work evaluates the performance of the assimilation of total lightning data within a three-dimensional variational (3DVAR) framework for the analysis and short-term forecast of the 24 May 2011 tornado outbreak using the Weather Research and Forecasting (WRF) Model at convection-allowing scales. Between the lifted condensation level and a fixed upper height, pseudo-observations for water vapor mass first are created based on either the flash extent densities derived from Oklahoma Lightning Mapping Array data or the lightning source densities derived from the Earth Networks pulse data, and then assimilated by the 3DVAR system. Assimilation of radar data with 3DVAR and a cloud analysis algorithm (RAD) also are performed as a baseline for comparison and in tandem with lightning to evaluate the added value of this lightning data assimilation (LDA) method.Given a scenario wherein the control experiment without radar or lightning data assimilation fails to accurately initiate and forecast the observed storms, the LDA and RAD yield comparable short-term forecast improvements. The RAD alone produces storms of similar strength to the observations during the first 30 min of forecast more rapidly than the LDA alone; however, the LDA is able to better depict individual supercellular features at 1-h forecast. When both the lightning and radar data are assimilated, the 30-min forecast showed noteworthy improvements over RAD in terms of the model’s ability to better resolve individual supercell structures and still maintained a 1-h forecast similar to that from the LDA. The results chiefly illustrate the potential value of assimilating total lightning data along with radar data.

Assimilation of Pseudo-GLM Data Using the Ensemble Kalman Filter

Total lightning observations that will be available from the GOES-R Geostationary Lightning Mapper (GLM) have the potential to be useful in the initialization of convection-resolving numerical weather models, particularly in areas where other types of convective-scale observations are sparse or nonexistent. This study used the ensemble Kalman filter (EnKF) to assimilate real-data pseudo-GLM flash extent density (FED) observations at convection-resolving scale for a nonsevere multicell storm case (6 June 2000) and a tornadic supercell case (8 May 2003). For each case, pseudo-GLM FED observations were generated from ground-based lightning mapping array data with a spacing approximately equal to the nadir pixel width of the GLM, and tests were done to examine different FED observation operators and the utility of temporally averaging observations to smooth rapid variations in flash rates. The best results were obtained when assimilating 1-min temporal resolution data using any of three observation operators that utilized graupel mass or graupel volume. Each of these three observation operators performed well for both the weak, disorganized convection of the multicell case and the much more intense convection of the supercell case. An observation operator using the noninductive charging rate performed poorly compared to the graupel mass and graupel volume operators, a result that appears likely to be due to the inability of the noninductive charging rate to account for advection of space charge after charge separation occurs.

Climatological analyses of LMA data with an open‐source lightning flash‐clustering algorithm

AbstractApproximately 63 million lightning flashes have been identified and analyzed from multiple years of Washington, D. C., northern Alabama, and northeast Colorado lightning mapping array (LMA) data using an open‐source flash‐clustering algorithm. LMA networks detect radiation produced by lightning breakdown processes, allowing for high‐resolution mapping of lightning flashes. Similar to other existing clustering algorithms, the algorithm described herein groups lightning‐produced radiation sources by space and time to estimate total flash counts and information about each detected flash. Various flash characteristics and their sensitivity to detection efficiency are investigated to elucidate biases in the algorithm, detail detection efficiencies of various LMAs, and guide future improvements. Furthermore, flash density values in each region are compared to corresponding satellite estimates. While total flash density values produced by the algorithm in Washington, D. C. (~20 flashes km−2 yr−1), and Alabama (~35 flashes km−2 yr−1) are within 50% of satellite estimates, LMA‐based estimates are approximately a factor of 3 larger (50 flashes km−2 yr−1) than satellite estimates in northeast Colorado. Accordingly, estimates of the ratio of in‐cloud to cloud‐to‐ground flashes near the LMA network (~20) are approximately a factor of 3 larger than satellite estimates in Colorado. These large differences between estimates may be related to the distinct environment conducive to intense convection, low‐altitude flashes, and unique charge structures in northeast Colorado.

Ground‐level observation of a terrestrial gamma ray flash initiated by a triggered lightning

AbstractWe report on a terrestrial gamma ray flash (TGF) that occurred on 15 August 2014 coincident with an altitude‐triggered lightning at the International Center for Lightning Research and Testing (ICLRT) in North Central Florida. The TGF was observed by a ground‐level network of gamma ray, close electric field, distant magnetic field, Lightning Mapping Array (LMA), optical, and radar measurements. Simultaneous gamma ray and LMA data indicate that the upward positive leader of the triggered lightning flash induced relativistic runaway electron avalanches when the leader tip was at about 3.5 km altitude, resulting in the observed TGF. Channel luminosity and electric field data show that there was an initial continuous current (ICC) pulse in the lightning channel to ground during the time of the TGF. Modeling of the observed ICC pulse electric fields measured at close range (100–200 m) indicates that the ICC pulse current had both a slow and fast component (full widths at half maximum of 235 μs and 59 μs) and that the fast component was more or less coincident with the TGF, suggesting a physical association between the relativistic runaway electron avalanches and the ICC pulse observed at ground. Our ICC pulse model reproduces moderately well the measured close electric fields at the ICLRT as well as three independent magnetic field measurements made about 250 km away. Radar and LMA data suggest that there was negative charge near the region in which the TGF was initiated.

Lightning channel length and flash energy determined from moments of the flash area distribution

AbstractA widely used approach in observational and modeling studies of NOx produced by lightning is to relate NOx production to the number of flashes, without regard for the distribution of lightning flash sizes. Recent studies have begun to consider channel length and flash size, which is now observable with VHF Lightning Mapping Array data. This study uses a capacitor model for flash energy based on the flash coverage area, which defines a size scale. This flash area is then filled with channel using a fractal method and compared to other methods that estimate length directly from the VHF source locations. In the presence of instrument measurement errors, area‐ and fractal‐based estimates are shown to be more stable estimators of flash length than connect‐the‐dots approaches and therefore are better suited for comparison to NOx production. A geometric interpretation of using vertical profiles of VHF source density to weight the altitude distribution of total channel length is developed. An example of the time series of moments of the lightning flash size distribution is shown for an example case, and some meteorological interpretation is given.

Environmental controls on storm intensity and charge structure in multiple regions of the continental United States

AbstractA database consisting of approximately 4000 storm observations has been objectively analyzed to determine environmental characteristics that produce high radar reflectivities above the freezing level, large total lightning flash rates on the order of 10 flashes per minute, and anomalous vertical charge structures (most notably, dominant midlevel positive charge). The storm database is drawn from four regions of the United States featuring distinct environments, each with coinciding Lightning Mapping Array (LMA) network data. LMAs are able to infer total lightning flash rates using flash clustering algorithms, such as the one implemented in this study. Results show that anomalous charge structures inferred from LMA data, significant lightning flash rates, and increased radar reflectivities above the freezing level tend to be associated with environments that have high cloud base heights (approximately 3 km above ground level) and large atmospheric instability, quantified by normalized convective available potential energy (NCAPE) near 0.2 m s−2. Additionally, we infer that aerosols may affect storm intensity. Maximum flash rates were observed in storms with attributed aerosol concentrations near 1000 cm−3, while total flash rates decrease when aerosol concentrations exceed 1500 cm−3, consistent with previous studies. However, this effect is more pronounced in regions where the NCAPE and cloud base height are low. The dearth of storms with estimated aerosol concentrations less than 700 cm−3 (approximately 1% of total sample) does not provide a complete depiction of aerosol invigoration.

On the relationship between continuing current and positive leader growth

AbstractIt has long been speculated that the source of continuing current (CC) for a negative cloud‐to‐ground flash is provided by the growth of its positive leader into negative charge regions. In this study, data from the Langmuir Electric Field Array (LEFA) and Lightning Mapping Array (LMA) are used to investigate these speculations. LEFA and LMA data provide a way to estimate the occurrence and duration of CC and channel growth throughout a flash, respectively. By connecting LMA VHF sources onto contiguous channels, the growth of the positive leader associated with each return stroke is inferred. A linear correlation between positive‐channel growth and CC duration is found, providing evidence that the positive leader grows with a constant velocity, but no obvious correlation of this velocity with CC occurrence is found. Each return stroke is then sorted by its channel growth rate and further identified by its CC type. This analysis also provides no identifiable correlation linking the positive‐channel growth rate to CC occurrence or duration. Finally, the positive‐channel growth rate for the whole flash is calculated in 10 ms windows so that any trends occurring before, during, or after the CC can be observed. This analysis too shows no correlation, which implies that positive‐channel growth is not the primary mechanism that determines CC occurrence and duration.

Theory and Observations of Controls on Lightning Flash Size Spectra

Abstract Previous analyses of very high frequency (VHF) Lightning Mapping Array (LMA) observations relative to the location of deep convective updrafts have noted a systematic pattern in flash characteristics. In and near strong updrafts, flashes tend to be smaller and more frequent, while flashes far from strong vertical drafts exhibit the opposite tendency. This study quantitatively tests these past anecdotal observations using LMA data for two supercell storms that occurred in Oklahoma in 2004. The data support a prediction from electrostatics that frequent breakdown and large flash extents are opposed. An energetic scaling that combines flash rate and flash area exhibits a power-law scaling regime on scales of a few kilometers and a maximum in flash energy at about 10 km. The spectral shape is surprisingly consistent across a range of moderate to large flash rates. The shape of this lightning flash energy spectrum is similar to that expected of turbulent kinetic energy spectra in thunderstorms. In line with the hypothesized role of convective motions as the generator of thunderstorm electrical energy, the correspondence between kinematic and electrical energy spectra suggests that advection of charge-bearing precipitation by the storm’s flow, including in turbulent eddies, couples the electrical and kinematic properties of a thunderstorm.

Lightning mapping observation of a terrestrial gamma‐ray flash

We report the observation with the North Alabama Lightning Mapping Array (LMA) related to a terrestrial gamma‐ray flash (TGF) detected by RHESSI on 26 July 2008. The LMA data explicitly show the TGF was produced during the initial development of a compact intracloud (IC) lightning flash between a negative charge region centered at about 8.5 km above sea level (−22°C temperature level) a higher positive region centered at 13 km, both confined to the convective core of an isolated storm in close proximity to the RHESSI footprint. After the occurrence of an LMA source with a high peak power (26 kW), the initial lightning evolution caused an unusually large IC current moment that became detectable 2 ms after the first LMA source and increased for another 2 ms, during which the burst of gamma‐rays was produced. This slowly building current moment was most likely associated with the upward leader progression, which produced an uncommonly large IC charge moment change (+90 C·km) in 3 ms while being punctuated by a sequence of fast discharge. These observations suggest that the leader development may be involved in the TGF production.

Estimations of charge transferred and energy released by lightning flashes

Data from ground‐based and balloon‐borne electric field meters are used in conjunction with lightning mapping array data to estimate the amount of charge transferred and energy liberated by 40 lightning flashes in four New Mexico mountain thunderstorms. Two models are developed to estimate the lightning charge transfer: point charges at the geometric center of two or more lightning charge distributions, and uniform charge distributions placed along the lightning channels. Results indicate that flashes with substantial horizontal extent, primarily intracloud (IC) flashes during the storm's mature stage, are better modeled with uniform charge distributions. Less extensive flashes, both cloud‐to‐ground (CG) and early‐stage IC flashes, are better modeled with point charges. The amount of charge transferred by 29 IC flashes ranged from ±3.9 to ±85.8 C, with an average of ±17.6 C. Mature‐stage IC flashes transfer significantly more charge, on average, than early‐stage IC flashes. The 11 CG flashes transferred negative charge to the ground; the average amount of charge lowered was −8.7 C. All but one CG flash also had substantial in‐cloud charge transfers (average ±9.3 C, maximum ±44 C) in addition to the charge moved to the ground. Individual lightning charge transfers are combined with concurrent in‐cloud potentials to estimate the electrical energy released by each flash. The IC flashes released an average of 1.6 GJ of energy per flash, while the average for CG flashes was about half as much energy per flash. Using average charge and potential values for the group of storms yields lightning energy estimates of 2 × 108 to 7 × 109 J, depending on flash type and storm stage.

Electrical evolution during the decay stage of New Mexico thunderstorms

The electric field (E) at the ground beneath a thunderstorm often exhibits an end‐of‐storm oscillation (EOSO) during the storm's decay phase in which E typically undergoes three polarity changes over a period of 30–75 min. To determine why the surface E oscillates, the timing of the EOSO polarity changes of a multicellular storm in New Mexico has been compared to a series of balloon, radar, and lightning data. Three in situ balloon soundings provide vertical profiles of E and inferred charge regions just before and during the EOSO. These data have been compared to the evolving precipitation structure determined from orthogonal range‐height scans of two radars and to several other parameters including lightning mapping array data. Before the EOSO began, the storm had the typical electrical structure of the late mature stage. Estimates of the total charge magnitudes in each of the principal regions before and during the EOSO ranged from 7 to 33 C. The comparisons indicate that surface E polarity changes during the EOSO were (1) primarily owing to the successive fallout of three of the four principal charge regions in the storm's mature stage charge structure, modified by (2) the growth or decay of the charge regions as they descended, and (3) changes in the screening charges at the upper and lower cloud boundaries. More limited data from two other decaying storms indicate that similar in‐cloud evolution was probably occurring during their EOSOs.